Population biologists and human geneticists are both interested in studies of genetic variability and the evolution of populations. This is because the change in frequencies of mutant genes is directly related to evolutionary processes as well as to human health care. Mathematical analysis is a powerful tool to study these problems and sophisticated theories have been developed. However, these theories generally ignore important non-genetic factors such as demography and sociocultural environments. Furthermore, recent developments in molecular biology enable us to study many genetic systems at the DNA level. Thus, the population dynamics of deleterious genes in human populations may be understood using data from a variety of disciplines such as molecular biology, psychology, sociology, anthropology, and population biology which seldom communicate with each other. For these purposes, I propose the following three projects: (1) Molecular population genetics in human populations. Molecular characteristics of many genetic systems have revealed new aspects of genome organization such as unequal crossing-over, gene conversion, and transposition of DNA segments. These mechanisms modify the expression of genes and change phenotypes drastically. Diseases such as thalassemias and diabetes and association between HLA haplotype and disease will be studied considering molecular characteristics of the genes involved. (2) Mechanisms for the maintenance of variation in the frequencies of deleterious genes in human isolates. It has been shown that the incidences of genetic diseases vary greatly among different ethnic groups. The study will provide one of the best opportunities to apply population genetics theories. (3) Population dynamics of deleterious genes under social selection. For several diseases, it has been shown that complex fitness interactions occur due to sociocultural reactions to affected individuals. Such fitness interactions are termed social selection and have been studied using mathematical models and existing data. The analyses show that disease incidence depends not only on the associated deleterious effect but also on sociocultural reactions to the disorder. Further analyses are necessary to extend our understanding of the phenomena of social selection.